Table 1_Longitudinal associations between PM2.5 with gestational diabetes mellitus mediated by gut microbiome and potential mechanism: based on a prospective pregnant women cohort in China.doc

BackgroundExposure to particulate matter pollution with aerodynamic diameters < 2.5 μm (PM2.5) has been linked to gestational diabetes mellitus (GDM) and gut microbiota dysbiosis. However, few studies have illustrated the associations among PM2.5 exposure, gut microbiota, blood metabolites, circular RNAs (circRNAs) and GDM risk. This study aimed to explore the moderating effects of the gut microbiota on the association between PM2.5 exposure and GDM, and to analyze the interaction network of PM2.5 exposure, gut microbiota, blood metabolites and circRNAs.
MethodsParticipants (n = 1,248) were selected from the Pregnancy Metabolic Disease and Adverse Pregnancy Outcome (PMDAPO) cohort in Guangzhou, China. Demographic information, blood and fecal samples were collected from the participants. The fecal microbial composition and relative abundance were characterized using 16S rRNA gene sequencing, while blood differential metabolites and circRNAs of pregnant women with GDM were assessed using non-targeted metabolomics and RT-qPCR, respectively. Exposure levels of air pollutants were assessed using data from the nearest monitoring station. Spearman correlation and regression models were conducted to estimate the associations among PM2.5 exposure, gut microbiota, blood metabolites, circRNAs and GDM.
ResultsElevated PM2.5 exposure levels were significantly associated with an increased risk of GDM, impaired glucose homeostasis and gut microbiota dysbiosis. Solobacterium and Escherichia_Shigella showed a positive effect modification on the association between PM2.5 exposure and fasting blood glucose, while Fusicatenibacter, Ruminococcaceae_UBA1819, Raoultibacter, Anaerofustis and Phascolarctobacterium showed a negative effect modification on the association between PM2.5 exposure and 2-h OGTT glucose. GDM-associated gut microbiota, including Catabacter, Angelakisella, Romboutsia and Fusicatenibacter, were associated with both GDM-associated metabolites (such as sphinganine-1-phosphate, sphingomyelin) and GDM-associated circRNAs (such as hsa_circ_0006732 and hsa_circ_0001439), which were involved in glycerophospholipid metabolism, sphingolipid metabolism and insulin signaling pathway.
ConclusionsThe gut microbiota may moderate the associations between PM2.5 exposure and blood glucose levels, and both PM2.5 exposure and gut microbiota may be related to GDM, potentially involving pathways such as glycerophospholipid metabolism, sphingolipid metabolism and the insulin signaling pathway. However, lifestyle factors (diet and physical activity) and residential mobility were not measured, and the fecal microbiota was assessed at a single time point in mid-pregnancy. Thus, these limitations may contribute to residual confounding, exposure misclassification, and limited causal inference.